Renders JavaScript Single Page Applications to extract LLM-ready Markdown content and capture screenshots.
Sponsored
SPA Reader addresses the challenge of extracting content from modern Single Page Applications (SPAs) that traditional web scrapers often fail to process due to their reliance on client-side JavaScript rendering. By leveraging a headless Chromium browser via Playwright, this tool simulates a user's browser, waits for pages to fully render, and then intelligently extracts clean, article-like Markdown using Mozilla's Readability and Turndown. The output is specifically formatted to be consumed by Large Language Models (LLMs), ensuring clean data for AI applications. It also provides capabilities for capturing high-quality screenshots of rendered pages.
주요 기능
01Includes robust SSRF protection and prevents Playwright selector injection
020 GitHub stars
03Truncates output content at 100KB with clean line-boundary cuts
04Renders SPA pages and extracts LLM-ready Markdown content (spa_read)
05Captures full-page or viewport-sized PNG screenshots (spa_screenshot)
06Reuses a single Chromium instance for efficient, low-overhead processing
사용 사례
01Providing clean, structured content from JavaScript-heavy websites to Large Language Models
02Generating visual representations (screenshots) of dynamically rendered web pages
03Integrating web content extraction capabilities into Model Context Protocol (MCP) clients
